Ellipsis Health is creating cutting-edge AI/ML products that solve healthcare staffing issues and administrative burdens using conversational AI and our patented voice biomarker technology in the delivery of better healthcare for everyone.
Requirements
- Experience with scripting languages. Also, experience working with or designing prompts for large language models (e.g., OpenAI GPT, Claude, PaLM, etc.)
- Strong writing and editing skills with an eye for tone, clarity, and conversational flow
- Background in healthcare, public health, medical writing, or conversational design for clinical settings is strongly preferred.
- Familiarity with HIPAA, PHI handling, and healthcare communication standards is a plus
- experience with scripting for voice interfaces, EHR integrations, or digital health applications
Responsibilities
- Design and evaluate prompts for AI agents and tools that support healthcare tasks such as patient outreach, enrollment, symptom screening, and follow-up care
- Work closely with clinical, product, and engineering teams to ensure prompt outputs align with medical accuracy, regulatory guidelines, and user intent
- Develop and maintain a prompt library tailored to different patient populations, clinical scenarios, and communication channels (voice, chat, etc.)
- Evaluate performance of LLM outputs, identify error patterns, and iterate using structured testing and evaluation
- Incorporate inclusive language, health literacy best practices, and cultural sensitivity into prompt design
- Stay informed on LLM capabilities, fine-tuning strategies, and ethical standards for AI in healthcare
Other
- Curious, self-starter, and a willingness to learn.
- Strong collaboration and communication skills with the ability to work across technical and clinical teams
- The typical salary range for this role is $80,000 to $150,000 USD.
- We offer competitive salary and benefits, including a 401k that matches up to a certain percentage of your salary, health, vision, and dental insurance, and very flexible paid time off.